Goal of repo: a) create voice model based on CMU ARCTIC data, and b) create a model based on an arbitrary set of vocabulary words, possibly arranged into sentences.
A script to automate training voice data for PocketSphinx. The only file you need is train_voice_model.py. Put this in the directory where you want your training data folder to be located. It will look for a folder called 'bespoke_training_data' and create it if necessary. Then you just have to sit there and read the text it feeds you on the screen. There's 500+ sentences to read, but you only really need about 20. I'd go for more.
You can pick up later if you need to; the script will look for the highest numbered file and start recording after that. I haven't tested what happens when it reaches the end of the list. Hopefully the exit is clean.